Robust control of DC microgrids

© 2017 Dr Majid FardRapid development in photovoltaic (PV) cells manufacturing technology and emerging environmental concerns have significantly increased photovoltaic penetration over the last decade. However, the inherent intermittency of solar radiation necessitates integration of other power generation sources to provide an uninterruptible, and reliable power supply that satisfies network requirements. Implementing energy storages can diminish the risk of power interruption due to PV system intermittency while ensuring demand satisfaction. To utilise PV and energy storage systems, a group of controllers should be implemented to control the PV systems to extract maximum power from each PV system, while the energy storage systems are employing as a secondary source of energy for a consistent operating of the whole network. In this thesis, initially, we will present a robust control design based on linear matrix inequality (LMI) for a small scale islanded (off-grid) DC hybrid system. This hybrid system is consisting of a photovoltaic (PV) system and an energy storage system (ESS). A robust H¥ controller is designed to regulate PV system input voltage such that the PV array is operating at maximum power point. Also, energy storage system is controlling to act as a complementary source of energy to maintain DC bus voltage at a constant value; This has achieved by designing a robust controller based on LMI with maximising the region of stability and minimising the RMS gain for the energy storage system. The design method considers both converters non-linearities, modelled as a convex polytope, and achieves H¥ performance. Secondly, a robust coupled controller is designed by taking into account the interaction between the PV and energy storage system components. Consequently, the proposed coupled controller causes a significant improvement in the system transient performance and efficiency. Furthermore, to expand the proposed approach for medium to large scale systems a new decentralised, autonomous power management structure for DC microgrids comprising of multi PV and energy storage units is proposed. Each unit is controlled by a robustly designed controller using its local measurements only. Thus, the requirements for communication links are omitted. Each control unit aims to deal with system parametric uncertainties and to guarantee a required disturbance rejection performance over both internal and external disturbances. Seven modes of operations are defined to cover all possible practical operating condition that can occur in a DC micro grid. The transition between the modes occurs autonomously and seamlessly for each control unit, and thus unit-to-unit communication is not required. Finally, to enable grid connection capability for the considered DC micro grid a new current control approach for three-phase grid-connected voltage source inverter (VSI) is proposed. An observer-based state-feedback dynamic controller is implemented to control the VSI. Moreover, the controller and observer are robustly designed, to account for grid disturbances, and achieve a guaranteed H¥ performance level. The existence of parametric uncertainties caused by inexact knowledge of grid impedance and inductance are modelled as a convex polytope. Additionally, by imposing a set of constraints that prevent the VSI from operating in the nonlinear region, saturation of control input is avoided, through overmodulation